CLiN. CHEM. 28/4, 955- 961 (1982)
Cytoskeletal Proteins from Human Skin Fibroblasts, Peripheral Blood
Leukocytes, and a Lymphoblastoid Cell Line Compared by TwoDimensional Gel Electrophoresis
1
Carol S. Giometti, Karen E. Willard, and N. Leigh Anderson
Differences in proteins between cells grown as suspension
cultures and those grown as attached cultures were studied
by comparing the proteins of detergent-resistant cytoskeletons prepared from peripheral blood leukocytes and
a Iymphoblastoid cell line (GM607) (both grown as suspension cultures) and those of human skin fibroblasts
(grown as attached cultures) by two-dimensional gel
electrophoresis. The major cytoskeletal proteins of the
leukocytes were also present in the protein pattern of
GM607 cytoskeletons. In contrast, the fibroblast cytoskeletal protein pattern contained four groups of proteins
that differed from the patterns of the leukocytes and
GM607. Three groups (Cytosk,:8- 10, : 14-16, and: 17 -18)
showed qualitative differences, and the fourth group (Cytosk,: 11 and: 13) showed quantitative differences. In addition, surface labeling of GM607 and human fibroblasts
with 1251demonstrated that substantial amounts of vimentin
and actin are exposed at the surface of the attached fibroblasts, but there is little evidence of similar exposure
at the surface of the suspension-grown GM607. Cytosk( 11
and: 13 in fibroblast preparations were also labeled with
the 1251. These results demonstrate some differences in
cytoskeletal protein composition between different types
of cells could be related to their ability or lack of ability to
grow as attached cells in tissue culture.
60000 (2). The major protein component of the intermediate
filaments in nonmuscle cells is vimentin (3). Actin is the major
protein in the microfilaments (4).
The distribution of cytoskeletal filaments within the cell
differs between cells grown in suspension and those that grow
as attached monolayers (1, 5), but whether there are corresponding differences in the protein composition of those
filaments in cells grown as suspensions or monolayers was
unknown . This paper compares the two-dimensional gelelectrophoretic patterns of cytoskeletal proteins from three
types of human cells. Two of the cell types-peripheral blood
leukocytes (primarily lymphocytes and monocytes) and the
lymphoblastoid cell line GM607-grow only as suspension
cultures. The third cell line- human skin fibroblasts-grows
in culture only as an attached monolayer.
The two-dimensional electrophoresis patterns show that
the fibroblasts contain several major cytoskeletal proteins that
are absent from both of the suspensipn-grown cell types. In
addition, surface labeling with 1251 shows that several of the
major fibroblast cytoskeletal proteins are exposed at the cell
surface, but little such labeling was obtained with the suspended cells. These findings provide a clue as to which unique
protein gene products associated with the cytoskeleton may
be involved with cell adhesion.
Materials and Methods
Radiolabeling of Cellular Proteins
Additional Keyphrases: surface vs interior cytoskeletal proteins . commonality of proteins in two sorts of cells • vimentin . actin • differences related to growth predilections
cell adhesion
The term "cytoskeleton" refers to the filamentous cytoplasmic structure that remains associated with nuclei after
extraction of whole cells with detergent solution. Three types
of filaments make up the cytoskeleton, two serving as structural components and a third functioning as a contractile
apparatus. In the former category are the microtubules, which
are the largest of the three filaments (25 nm in diameter), and
the intermediate filaments (10 nm in diameter) (1). Microtubules appear to be involved with chromosome movement
during mitosis, intracellular transport of organelles, the
movement of cells during tissue differentiation, and the formation and maintenance of cellular processes. The function
of the intermediate-size filaments remains obscure, but some
evidence suggests that these structures may also be involved
with organelle movements. The smallest of the three cytoskeletal elements, the microfilaments (6 nm in diameter), are
directly involved in cytoplasmic streaming, cytokinesis,
membrane ruffling, cell locomotion , endocytosis, and exocytosis (J). The major protein component of the microtubules
is tubulin, a complex of three proteins (0'1,0'2, and (3 tubulins)
with relative molecular masses (Mr) between 55000 and
Molecular Anatomy Program, Division of Biological and Medical
Research, Argonne National Laboratory, Argonne, IL 60439.
Human peripheral blood leukocytes were isolated as previously described (6). The leukocytes (4 X 106 cells/mL) were
cultured for 18- 24 h in flat-bottomed multi-well plates containing methionine-free RPMI 1640 medium (GIBCO, Grand
Island, NY 14672) with added fetal bovine serum (50 mL/L) ,
2-mercaptoethanol (4 X 10- 5 mol/L), and 60 mCi/L of [358 ]methionine (Amersham, Arlington Heights, IL 60005).
Cells from the human lymphoblastoid cell line GM607
(Human Genetic Mutant Cell Repository, Institute for
Medical Research , Camden, NJ) were labeled with [35 8]methionine in an identical manner.
Human skin fibroblasts (cell line 1494 from Meloy Laboratories, Springfield, VA; National Cancer Institute Contract
N01-CP91000) were cultured (5 X 104 cells/well) for 24 h in
flat-bottomed multi-well plates containing Eagle Minimal
Essential Medium (MEM; GIBCO) with 100 mL of added fetal
bovine serum per liter, in order to obtain semiconfluent cultures. The medium was then removed, replaced with methionine-free ¥EM containing fetal bovine serum (50 mL/L) and
[ 35 8]methionine (60 mCi/L), and the cells were cultured for
an additional 18- 24 h.
Cell-surface proteins were labeled with 1251 by using the
glucose oxidase- lactoperoxidase reaction as described by
Hynes (7). Human fibroblasts were iodinated while they were
attached to the tissue-culture plates, but GM607 cells were
iodinated as a cell suspension.
Preparation of Cytoskeletons
Cytoskeletons were prepared from peripheral blood leuCLINICAL CHEMISTRY, Vol. 28, No.4, 1982
955
kocytes and GM607 by resuspending washed cell pellets
(approximately 8 X 106 cells) in 5 mL of phosphate-buffered
isotonic saline (without calcium and magnesium; GIBeo)
containing, per liter, 180 mL of glycerol, 5 mL of Nonidet P40
detergent (NP40; Particle Data Laboratory Ltd. , Elmhurst,
IL 60126), 1 mmol of phenylmethylsulfonyl fluoride, and 875
/-Lmol of pepstatin (a hexapeptide carboxyprotease inhibitor).
After incubation at room temperature for 10 min, the samples
were centrifuged for 15 min at 700 X g (JS 3.0 rotor, Beckman
J-6B centrifuge). The supernates were discarded and the
pellets were solubilized as described below.
Fibroblast cytoskeletons were prepared as described previously (8).
Sample Preparation for Electrophoresis
Labeled cells and cytoskeleton preparations were harvested
for electrophoresis with 50 /-LL/well of a solution containing,
per liter, 9 mol of urea, 50 mL of NP40, 20 g of LKB ampholyte
(pH range 3.5- 10; LKB Instruments, Inc., Rockville , MD
20852),50 mL of 2-mercaptoethanol , and 1.0 mmol of ph enylmethylsu lfonyl fluoride. Samples were centrifuged for 30
s in a Beckman Microfuge B to remove insoluble material, and
20 /-LL of t he resulting supernates was analyzed by two-dimensional gel electrophoresis.
Two-Dimensional Gel Electrophoresis
First-dimension isoelectric focusing was done by using the
ISO apparatus (9) and LKB ampholytes (10% pH range, 2.5-4;
90% pH range, 3.5-10). The proteins were foc used for 14 000
V - h at room temperature. The second-dimension separation
(sodium dodecyl sulfate electrophoresis) was done on linear
10-20% polyacrylamide gradient gels as described (10). Gel
pouring, electrophoresis, staining, and destaining were all as
described previously (J 1). Gels of whole-cell proteins were
autoradiographed (J 2); gels of cytoskeletal proteins were
fluorographed (13) .
Results
Figure 1 shows the two-dimensional electrophoresis patterns of [ ~5S1m et hi on in e- lab e l ed proteins from the three
human cell types studied. The intermediate fi lament protein,
vimentin, is located in an id entical position in all three patterns, as is actin. The IgM heavy-chain and light-chain proteins, distinctive markers of the lymphoblastoid cell line
GM607, are prominent only in the GM607 pattern. The areas
out lined in the three patterns of Figure 1 are proteins known
to be associated with the detergent-resistant cytoskeletons of a ll three cell types. Although the protein patterns of
the leukocytes (A) and GM607 (B) are identical in the regions
outlined, the protein pattern of the fibroblasts (C) has very
different constellations of spots in three of the four areas
outlined.
Preparation of detergent-resistant cytoskeletons from cell
suspensions requires more manipulation tha n does similar
preparation from attached cells. Fibroblast cytoskeletons
remain attached to the plastic tissue-culture dish after detergent extraction and can be easily removed with the
NP40/urea solubilization mixture described above. However,
cytoskeletons from peripheral blood leukocytes and GM607,
which grow as suspension cultures, must be pelleted by centrifugation after detergent extract ion . Because the cytoskeleton is still associated with the nucleus under the extraction
conditions described, a short low-speed centrifugation suffices
to collect the cytoskeletons. Figure 2 shows that this technique
yields cytoskeletons of peripheral blood leukocytes (A) that
are as pure as those prepared from attached fibroblasts. The
GM607 cytoskeleton preparations (B) consistently contain
956
CLINICAL CHEMISTRY. Vol. 28. No.4. 1982
more protein components t han do the fresh leukocyte preparations, despite their simi lar growth conditions. Comparison
with Figure IB, however, shows t hat even t he GM607 cytoskeleton preparations are enriched in specific proteins. The
extra proteins seen in the GM607 cytoskeletal preparations
may be actual cytoskeleta l components peculiar to the longterm culture conditions of this cell line, or they may indicate
that the lymphoblastoid cells are more resistant to detergent
extraction, requiring longer extraction times or higher concentrations of detergent. The appearance of the IgM heavychain and light-chain proteins with the cytoskeletal proteins
indicates that some membrane components are still associated
with the cytoskeleton in these GM607 preparations.
As mentioned, intact nuclei are associated with the. cytoskeletal filaments after detergent extraction. Two major
nuclear proteins (identified by co-migration with proteins
from purified nuclei) are visible in the patterns of cytoskeletal
proteins from peripheral blood leukocytes and GM607.
Comparable proteins have not as yet been identified in the
fibroblast pattern. The heat-shock protein (J 4) is found associated to some extent with the cytoskeletons of all three cell
types. Vimentin and actin are present with almost equal
abundance in cytoskeleton preparations from all three types
of cells, although a bit more vimentin appears to be associated
with fibroblast cytoskeletons than with t hose of either of the
suspension-cell types. The presumed degradation products
of vim en tin (Cytosk:1 and :4- 7) are not visible in the GM607
pattern, a find ing specifi c to this cell line.
The major cytoskeletal proteins that have not yet been
identified are designated by numbers (8, 15). Of these, Cytoskr:l1 and :13 (subscript f signifyi ng fibroblast proteins) are
beli eved to be identica l to Cytosk):l1 and :13 (subscript I signifying leukocyte proteins) because they co-migrate electrophoretically. However, the leukocyte and lymphoblastoid
cytoskeleton preparations always show a greater abundance
of Cytosk:13 t han Cytosk:ll, while the reverse is true of fibroblast cytoskeleton preparations. Cytoskr:ll, but not :13,
has been found to copurify with nonmuscle tropomyosin (8).
In contrast, preliminary data show that both Cytoskd1 and
:13 copurify with nonmuscle tropomyosin, although Cytosk):l1
is more abundant (i.e., more enriched ) in such preparations
than Cytosk ):13.
The fibroblast cytos keletal proteins Cytoskf:8-10 have no
counterparts in the leukocyte or lymphoblastoid cytoskeleton
preparations. Cytosk):8, thought to be a possible candidate
for a protein analogous to Cytoskr:8, has a completely different
peptide pattern after partial proteolysis with chymotrypsin
(data not shown). Therefore, Cytoskr:8-10 appear to be unique
to fibroblasts, and possibly to attached cell s [with the exception of peripheral blood monocytes (16)], while Cytosk):8 is
unique to leukocytes, or possibly to cells that grow as suspension cultures.
The fibroblast proteins Cytosk(14-18 are enriched, together
with actin, in nonmuscle myosin preparations done according
to Burridge and Bray (data not shown) (17). In the whole-cell
protein pattern (Figure 1C), Cytoskr:14, :15, and :16 each have
two "satellite" spots slightly above and to the left of (i.e., acidic
to) t he major protein spot. These "satellite" spots are phosphorylated forms of Cytoskr:14, :15, and :16, analogous to the
phosphorylated myosin light chains of muscle (18). The periphera l blood leukocyte and lymp hoblastoid cytoskeletons
show on ly three proteins in positions simi lar to the myosin
light chains of fibroblasts, and these have been numbered to
correspond to the fibrob last proteins based on their co-migration characteristics- i.e., Cytosk):1 5, :16, and :18. Cytosh15 and :16 each have phosphorylated forms that can be
seen in the whole-cell patterns (Figure 1,A and B), suggesting
that they too may be myosin light chains modified by phosphorylation.
A
c
8
•
.1r
1-
-
-:';-
.. ~I;- "'·..... •
G]- ......
~ .-
•.
...
-
..
o
r
Z
o
}>
r
o
[;:]
"
.
•
..
....
~
..
- . ..-..
-",.- .. ..,.
. ~..
"
D
~
.;
•
cs5740c
"t
:: ;
.
..
•
~
...
..
.,
m
Cii
~
-<
<
~
I\>
.OJ
z
!='
•.f>.
.....
<0
OJ
I\>
ID
U1
.....
-.
"'~/(LC
-"
-
0
~
...
cs5081
Fig _ 1. Two-dimensional electrophoresis patterns of total cellular proteins from cells grown as suspension cultures (A and B) or
as attached mono layers (C)
A, peripheral blood leukocytes; B, Iymphoblastoid cell line GM607; C, human skin fibroblast line 1494. Abbreviations: V, vimentin; A, actin , IgM·HC, IgM heavy
chain; KLC, kappa light chain. Spot constellations inside rectangles differ between cytoskeleton preparations from suspension culture cells and attached monolayer
fibroblasts. Spot constellation within the circles is found in cytoskeletons from all three types of cells. Patterns are oriented with the acidic side to the left, basic
side to the right
•
~ ~
~;
•
I
s::
..
-
"'!'
.t.
..
..
-
...-
•
D
.-
........
-...
-
• •.
•
~
'
•
!'• ..,:
...
. . -- ....
..
# '
~~.
.r_
cs 5760b
CD
<II
CD
()
r
Z
0
»
r
()
-
I
m
s::
Cii
--t
:IJ
--<
A
<
c
B
~
.
I\)
~
z
N, :1
!'-
.1\
-..
!=>
v~c,:;
~
<0
ex>
I\)
C,:1--.. I.
~-:
C, :6-. \
C, :7- • C, :5
.
Nt l
{\ '
.'
··'HS:'
.
--
· -~~S : '
"
'HS:I
•
•
:+,-:2 . 'N,:2
~
A
..
C, :8C, :8-
'CI :13
C, :11/
(nmTm)
C, :11/ - 'C,:13
'\.
,.
.
.
C(S. --C(9
.
·--CnD
Cfl l'"
'C(l3
(nmTm)
'KLC
(nmTm)
C, :15
-
C, :15-. C,:16
-
C('5
I
CI :16
C('4
..... C('6
C(IS
\
'CI :18
-Cft?
'C,:18
cs4911
cs 5091
Fig. 2. Two-dimensiona' electrophoresis patterns of cytoskeletal proteins from cells grown as suspension cultures or as attached
monolayers
A, peripheral blood leukocytes; B, Iymphoblastoid cell line GM607 ; C, human skin fibroblast line 1494, Abbreviations as in Figure 1, with these additions: N, nuclear
protein; HS, heat-shock protein; C, cytoskeletal protein , nmTm, non muscle tropomyosin . Subscripts (lor f) refer to proteins found in leukocytes (including GM607)
or fibroblasts, respectively, Patterns oriented as in Figure 1
cs5632
A
..
..
cs5720a
Fig. 3. Two-dimensional electrophoresis patterns of 12sl-labeled proteins from whole cells
A, GM607; B, human fibroblasts. Abbreviations as in Figures 1 and 2, with these additions: HLA, human histocompatibility antigen HL-A; 132M, {32-microglobulin .
Patterns oriented as in Figure 1
Reportedly, nonmuscle contractile proteins known to be
components of the cytoskeleton are sometimes found on the
surface of cells (19, 20). To compare the possible surface expression of cytoskeletal proteins in cells grown as suspensions
versus those grown as monolayers, we labeled whole GM607
and human fibroblasts with 1251 by the glucose oxidase-lactoperoxidase reaction (7). Heavy labeling of actin and vimentin was detected in the two-dimensional electrophoresis
patterns of the iodinated fibroblast proteins, and Cytoskr:11
and :13 were faintly visible (Figure 3). In contrast, actin could
barely be seen in the pattern from iodinated GM607 proteins
(Figure 3A), and it was difficult to identify which, if any, of
the smudges above and to the left of actin might be vimentin.
This difference in labeling efficiency of actin and vimentin in
the two cell lines shows that the iodination was restricted to
t he surface of the cells, because labeling of internal proteins
would have resulted in equal labeling of actin and vimentin
in both cell types. The labeling of the HL-A antigen and associated !J2-microglobulin in GM607 preparations further
demonstrated the surface specificity of the label. These data
suggest that the microfilaments and intermediate filaments
are more exposed at the cell surface of attached cells than is
the case for cells in suspension cultures.
Discussion
In this preliminary study, the cells used for analysis by
two-dimensional electrophoresis were from different individuals and were grown under different culture conditions.
In addition, the growth characteristics of the cells used-Le"
growth as attached or suspension cultures- necessitated the
use of two different methods of cytoskeleton preparation.
Could any or all of these differences in experimental condi tions explain the cytoskeletal protein variations observed
when the two-dimensional electrophoresis patterns of human
leukocytes, GM607, and skin fibroblasts were compared? We
have analyzed skin fibroblasts from over 100 different humans
and in every case have found all of the major cytoskeletal
proteins discussed in this paper to be located in positions
identical to those shown in Figure 2C, whether the cells were
grown in RPMI 1640 (8) or MEM. In cases where variants of
cytoskeletal proteins have been found, the variant protein has
been present in addition to those normally seen in fibroblast
patterns (8). In addition, analysis ofleukocyte samples (both
normal and leukemic) from over 100 different humans and ten
different lymphoblastoid cell lines has demonstrated that the
proteins described as fibroblast-specific cytoskeletal proteins
(e.g., Cytoskf:8) are never found in leukocytes, and the leukocyte-specific cytoskeletal proteins (e.g., Cytoskl:8) are never
found in fibroblasts. Therefore, the possibility that the differences shown in Figures 1 and 2 are due to individual polymorphisms or culture conditions is unlikely. In experiments
where we attached leukocytes or GM607 to culture dishes by
using concanavalin A and subsequently prepared cytoskeletons according to the methods described for fibroblasts,
two-dimensional electrophoresis patterns of the cytoskeletal
proteins were identical to those shown in Figure 2A and B with
the addition of some cell-surface proteins probably involved
with binding to the concanavalin A. None of the human fibroblast-specific proteins were found. Therefore, the preparation of cytoskeletons from cells that are in suspension or
attached has no effect on the protein population that is recovered. As discussed below, the comparison of proteins
present in suspension and attached cells will be made more
complete by the future analysis of a cell line that can be
stimulated to convert from one type of culture to the other
(21). However, the protein differences reported here are true
variations among the three cell types analyzed and should be
considered as candidate proteins for involvement with cell
attachment.
The appearance of differences in cytoskeletal protein
composition between cells grown as suspension cultures and
those grown as attached monolayers is not surprising, because
CLINICAL CHEMISTRY, Vol. 28, No.4, 1982
959
the cytoskeletal fil aments are thought to be responsible for
cell shape and, perhaps, attachment sites. The interesting
problem is to determine in which of the cytoskeletal proteins
the expected heterogeneity is found.
Our comparison of two types of human cells grown as suspension cultures showed that all of the cytoskeletal proteins
found in peripheral blood leukocytes are also found in lymphoblastoid celis, except for those proteins believed to be
vimentin degradation products (Cytoskd and :4- 7). This
differe nce suggests the absence of a proteolytic enzyme in the
Iymphoblastoid line that is present in all other cell types analyzed thus far, both attached and unattached, or indicates
a difference in the GM607 vimentin structure that makes it
more resistant to proteolysis than is the vimentin found in
other types of cells. There are proteins associated with t he
Iymphoblastoid cytoskeletons t hat are not present in the
pattern for the peripheral blood leukocyte. These proteins
may be cytoskeletal components unique to lymphocytic celis
grown in long-term cultures, because the same proteins can
be found in the two-dimensional electrophoresis pattern of
total proteins from GM607 but not the pattern of peripheral
blood leukocytes (predominantly lymphocytes). Another
possibility is that the lymphoblastoid cells contain many
noncytoskeletal proteins that adhere to the cytoskeletons as
prepared for these studies. Only the cytoskeletal proteins
common to both the peripheral blood leukocytes and the
GM607, but different from the major cytoskeletal proteins of
the fibroblast, can be considered candidates for suspensioncell -specific proteins.
Comparison of the cytoskeletal protein patterns from leukocyte and GM607 with patterns from similar preparations
of human fibroblasts showed several groups of proteins that
were unique to the fibroblast. The differences seen between
the myosin light-chain regions (Cytoskr:14-18) of the human
fibroblast and suspension-culture patterns, however, have also
been found in a comparison of proteins from the fibroblasts
of different animals. For example, cytoskeletons from fibroblasts of mink and dog lack Cytoskr:14, as do human leukocytes and GM607 (data not shown). Therefore, the protein
heterogeneity found in this region of the patterns reflects some
functional difference between cells other than simply the
characteristic of attachment or non attachment in tissue culture. In contrast, the expression of Cytoskr:8- 10, although t he
number of proteins in this set varies among animals, has been
seen only in attached cell cultures, suggesting that those
proteins are required for some aspect of fibroblast structure
or function. 125I-surface l!lbeling of fibroblasts showed no
evidence of either the Cytoskr:8-10 or Cytoskr:14-18 spot
constellations, indicating that these proteins are not externally
involved with cell attachment.
Of the groups of cytoskeletal proteins that differ between
human fibroblasts and leukocytes (including GM607), the
Cytosk:ll and :13 pair is the most well-characterized. Both
proteins are present in all three cell lines studied, but the ratio
of Cytosk: 11 to :13 appears to be related in some way to cell
attachment or nonattachment. An interspecies comparison
of fi broblast cytoskeletal proteins has shown that Cytoskr: 11
is always more abundant that Cytoskr:13 in fibroblasts from
species as diverse as cows and dolphins (data not shown) . Both
the human peripheral blood leukocytes and GM607, however,
contain more Cytoskl:1 3 than Cytosk1:l1. Similar results have
been obtained from the analysis of purified populations of
human granulocytes and lymphocytes from peripheral blood
(16). In each case, Cytosk1:13 is more abundant than Cytoskl:l1. However, monocytes, after several days in culture,
will grow as attached cells and have almost equal amounts of
Cytosk1:ll and :13 (16) . This variation in abundance and the
copurification of both Cytosk1:ll and :13 with nonmuscle
tropomyosin fro m leukocytes and GM607 but only Cytoskr:ll
960
CLINICAL CHEMISTRY, Vol. 28, No. 4, 1982
from fibroblasts poses interesting questions as to the number
of tropomyosin genes present in the hum an genome and the
regulation of their expression in different cell types. For study
of the expression of proteins involved with cell attachment,
cytos keletal proteins from a cell line such as HL60 (21), which
can be converted from a suspension culture to an attached
culture at various points in the cell-alteration process, should
be isolated and electrophoresed.
Other proteins in the cytoskeletal preparations that thus
far are unidentified may also be found to have significance in
the comparison of attached with nonattached cells. These may
be additional nuclear proteins, proteins associated with
polyribosomes (22), proteins serving regulatory functions such
as the microtubule-associated proteins (23), or membrane
proteins that are tightly associated with the cytoskeleton in
a manner similar to the 5' -nucleotidase (EC 3. 1.3.5) t hat is
attached to the detergent-resistant matrix of murine lymphoid
cells (24). The microtubule proteins remain to be compared,
because tubulin and many associated proteins were extracted
from the cytoskeletons under the conditions we used for these
experiments (23) .
Our data demonstrate that the cytoskeletal protein composition of cells is tailored to specific cell structures and
functions. Such data, essential for the Human Protein Index
(25), allow us to begin to catalog those proteins t hat are
common to all human cell types, those that are specific to
certain cells, and those that vary during different stages of
normal cell differentiation or pathological processes. Recently,
Bachvaroff et al. used two-dimensional electrophoresis to
show that the cytoskeletal proteins actin and tubulin could
be detected on the surface of human leukemia cells and of
human lymphocytes transformed by mitogens or EpsteinBarr virus, but not on normal resting lymphocytes (18). The
identification of other cytoskeletal proteins, such as nonmuscle tropomyosin and myosin, on two-dimensional electrophoresis patterns of human cells and knowledge of their
normal occurrence could extend such studies and perhaps
provide specific clinical markers for alterations in cell structure that correlate with diseases such as leukemia.
We thank Dr. William Blattner for his assistan ce in obtaining
human sk in fibroblast line 1494 and Anne Gemmel for her excellent
techni cal help with cell cultures. This work was supported by the U.S.
Dept. of Energy und er contract no. W-31 -109-ENG -38.
References
1. Goldman, R. D., and Knipe, D. M., Functions of cytoplasmic fibers
in non -muscle cell motility. Co ld Spring Harbor Symp. Quant . Bioi.
37,523-534 (1973) .
2. EI-Hamalawi, A. A., Isolation and separat ion of (l'- and ,B-tubulin
from chick-embryo brain, muscle, and skin. Biochem. J. 169, 717- 719
(1978).
3. Granger, B. L., and Lazarides, E., Desmin and vimentin coexist
at the periphery of the myofibril Z disc. Ce ll 18,1053-1063 (1979).
4. Schloss, J. A., a nd Goldman, R. D., Microfilaments and tropomyosin of cultured mammalian cells: Isolation an d character ization. J.
Cell Bioi. 87,633- 642 (1980) .
5. Sundq vist, K., Otteskog, P., Wanger, L., et ai., Morphology and
microfi lament organi zation in hum an blood lym phocytes. Effects of
substratum and mitogen exposure. Exp. Cell Res. 130, 327-337
(1980).
6. Willard, K. E.. and Anderson, N. G., Two-dimensional analysis of
human lymphocyte proteins: I. An assay for lym phocyte effectors.
Clin. Chem. 27,1327- 1334 (1981) .
7. Hynes, R. 0., Alteration of cell-surface proteins by viral transformation and by proteolysis. Proc. Natl. Acad. Sci. USA 70, 3170-3174
(1973).
8. Giometti, C. S., and Anderson, N. L., A vari ant of hum an nonmuscle tropomyos in found in fibroblasts by using two-dim ensional
electrophoresis. J. BioI. Chem. 256,11840- 11846 (1981).
9. Anderson, N. G., and Anderson, N. L. , Analytical techniques for
cell fractions. XX I. Two-dimensional analysis of serum and tissue
protei ns: Multip le isoelectri c focu sing. Anal. Biochern.. 85,331- 340
(1978) .
10. Anderson, N. L. , and Anderson, N. G., Ana lytical techniqu es for
cell fr actions. XXII. Two-dimensional analysis of se rum and tissue
proteins: Multiple gradient slab-gel electrophoresis. Anal. Biochern..
85, 341-354 (1978).
11. Anderson, N. L., and Anderson, N . G., High resolution twodimensional electrophoresis of human plasma proteins. Proc. Natl.
Acad. S ci. USA 74,5421- 5425 (1977).
12. Willard, K. E. , and Anderson, N. L., Alterations of two-dimensional electrophoretic maps in human peripheral blood lymphocytes
induced by concanavali n A. In Electrophoresis '79, B. J. Radola, Ed.,
Walter de Gruyter, New York, NY, 1980, pp 413- 424.
13. Bonner, W. M., and Laskey, R. A., A film detection method for
tritium-labeled proteins and nucleic acids in polyacrylamide gels. Eur.
J . Biochern.. 46,83- 88 (1974).
14. Anderson, N. L., Gemmel, M. A., Giometti, C. S., and Anderson,
N. G., An investigation of heat shock-induced proteins of human cells
using two-dimensional electrophoresis. Clin. Chern. . 28, 1084- 1092
(1982).
15. Anderson, N. L., Identification of mitochondrial proteins and
some of their precursors in two-dimensional electrophoretic maps of
human cells. Proc . Natl. Acad. Sci. USA 78,2407- 2411 (1981).
16. Gemmel, M. A., and Anderson, N. L., Lymphocyte, monocyte,
and granulocyte proteins compared by use of two-dimensional electrophoresis. Clin. Chern.. 28, 1062- 1066 (1982) .
17. Burridge, K. , and Bray, D. , Purification and structural analys is
of myosins from brain and other non-muscle tissues. J. Mol . Bioi. 99,
1- 14 (1975).
18. Giometti, C. S., Anderson, N . G. , and Anderson, N . L. , Muscle
protein analysis. 1. High- resolution two-dimensional electrop horesis
of skeletal muscle proteins for analysis of small biopsy samples. Clin.
Chern.. 25,1877- 1884 (1979).
19. Bachvaroff, R. J., Miller, F., and Rapaport, F . T ., Appearance of
cytoskeletal components on the surface of leukemia cells and of
lymphocytes transformed by mitogens and Epstein- Barr virus. Proc.
Natl . Acad. Sci. USA 77,4979- 4983 (1980).
20. Owen, M. J., Auger, J. , Barber, B. H., et aI., Actin may be present
on the lymphocyte surface. Proc. Natl. Acad. Sci. USA 75,4484-4488
(1978).
21. Huberman, E., and Callaham, M. F., Induction of terminal differentiation in human promyelocytic leukemia cells by tumor-promoting agents. Proc. Natl. A cad. Sci. USA 76,1293- 1297 (1979).
22. Lenk, R., Ransom, L., Kaufmann, Y. , and Penman, S., A cytoskeletal structure with associated polyribospmes obtained from
HeLa cells. Cell 10,67- 78 (1977) .
23. Solomon, F., Magendantz, M ., and Salzman, A. , Identification
with cellular microtubules of one of the co-assembling microtubuleassociated proteins. Cell 18,431- 438 (1979) .
24. Mescher, M . F., Jose, M. J. L., and Balk , S. P., Actin-containing
matrix associated with t he plasma membrane of murine tumor and
lymphoid cells. Nat ure 289, 139-144 (1981).
25. Anderson, N. G., and Anderson, L., Automatic chemistry and the
Human Protein Index. J. Autorn.. Chern.. 2,177- 178 (1980).
CLINICAL CHEMISTRY, Vol. 28, No.4, 1982
961